JPH11168455A - Digital pll circuit and signal reproducing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a digital PLL circuit and a signal reproducing method with which pull-in time can be shortened while efficiently utilizing a data area without lowering durability to the jitter fluctuation or duty distortion of an input data signal. SOLUTION: This circuit has a data sampling processing part 1 for sampling the input data signal and outputting N pieces of sample data signals, data reproducing processing part 3 for outputting a regenerative data signal based on the sample data signal 6 outputted from the data sampling processing part 1 and an N-phase clock signal and delay processing part 2 for delaying the inputted sample data signals 6 and outputting them to the data reproducing processing part 3 as delay sample data signals 7, and time to the selective output of the extracted clock signal based on the phase information of the input data signal is shortened rather than time to the output of the input data signal as the regenerative data signal.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a digital PLL.
(Phase Locked Loop, hereafter simply PL
Indicated as L. The present invention relates to a circuit and a signal reproducing method, and particularly to a digital PLL circuit and a signal reproducing method used in an optical communication system having a passive double star structure or the like.

[0002]

2. Description of the Related Art Conventionally, information communication technology has expanded, and the need to transmit a large amount of data at high speed has arisen. Accordingly, the Institute of Electronics, Information and Communication Engineers 1997 Fall Society Conference C-12-25, C-12-26, The Institute of Electronics, Information and Communication Engineers 1996 Fall Society Conference SC-13-5,
As represented by B-844 and the like, a digital PLL circuit for extracting a clock signal from a burst-like data signal at high speed to reproduce data and a signal reproducing method associated therewith have been actively studied. ing.

In general, such a high-speed digital PL
In order to realize an L circuit, a pull-in operation of outputting a reproduced data signal and an extracted clock signal from a burst-like data signal at a high speed of several bits is required.

However, the term “pull-in” used herein means an operation of extracting a reproduced data signal having no identification error from an input burst data signal.

When the input burst data signal is numbered in order of 1, 2, 3,... One bit at a time from the beginning, a pull-in time corresponds to the number assigned in that order. It is used in the sense of the number of bits from which bit of the data signal the identification error is eliminated.
Therefore, the “pull-in time” here has no relation to the delay time required from the input of the input data signal to the output of the reproduced data signal.

The above-mentioned "pull-in" and "pull-in time" will be described with reference to FIG.

FIG. 13 shows a conceptual diagram of an input data signal transmitted from a terminal and a conceptual diagram of a reproduced data signal obtained by reproducing the transmitted data.

As shown in FIG. 13A, a bit number is assigned to each bit of the data area of the input data signal from the beginning. Then, as shown in FIG. 13B, in the reproduced data signal, data having no identification error is reproduced from the third bit.

Here, the overhead added to the data area is a training bit for performing clock recovery or the like.

Therefore, in the case shown in FIG. 13, it can be seen that the pull-in time is 3 bits. In the following description, “pull-in” and “pull-in time”, which are important concepts in the present invention, are used in the above-mentioned meaning.

Next, a configuration of a conventional digital PLL circuit and a signal reproducing method will be described with reference to FIG. FIG. 16 shows a block diagram of a configuration of a conventional digital PLL circuit.

As shown in FIG. 16, a conventional digital PLL circuit receives an input data signal and an N-phase clock signal, and samples the input data signal using the N-phase clock signal to obtain N samples. It has a data sampling processor 1 that outputs a sample data signal 6 composed of a data signal.

Further, the value of each of the input sample data signals 6 is fetched by the extracted clock signal, and the value indicating the phase number of the clock signal indicating the position of the rising change point of the input data signal and the value of the falling change point are indicated. A value indicating the phase number of the clock signal indicating the position is calculated, an average value of these values is calculated, and the number of rising and falling transition points in one period of the extracted clock signal is calculated. Is calculated and output as a change point calculation output signal 8.

Further, a value indicating the average value of the phase number of the rising change point represented by the change point calculation output signal 8 and the average value of the phase number of the falling change point output from the change point detection calculation section 4 And an extracted clock selection processing unit 5 that outputs an extracted clock based on one of the following values:

Further, the sample data signal 6 output from the data sampling processing section 1, the change point calculation output signal 8 output from the change point detection calculation section 4, and the extracted clock output from the extraction clock selection processing section 5. And a data reproduction processing unit 3 for outputting a reproduction data signal synchronized with the extracted clock signal based on the signal.

A digital PLL circuit shown in FIG.
In the signal reproducing method using the digital PLL circuit, the input data signal is sampled by an N-phase clock signal to obtain N sample data signals, and the sampled data signal is sampled by a predetermined clock to thereby obtain an input signal. The information of the change point of the data signal is detected, and based on the detected information, the extracted clock signal is
It selects from the N-phase clock signal and outputs a reproduced data signal based on the selected extracted clock signal.

A conventional digital PLL having such a configuration
The circuit and the signal reproducing method realize single-core bidirectional optical communication using an optical fiber in an optical communication system of a passive double star structure in which terminals and a transmitting / receiving station are connected in a star shape by a star coupler or the like, for example. Often used for.

On the other hand, in an optical communication system, as a structure of data exchanged between a transmitting / receiving station and a terminal, for example, as shown in FIG. Normally, the area is transmitted, and the terminal transmits a data area including overhead synchronized with the clock of the transmitting / receiving station.

Data transmitted from a terminal or a transmitting / receiving station causes duty distortion, jitter, frequency deviation, etc. due to the influence of an optical transmission path length, a transmitting / receiving circuit, and the like. What responds to such deterioration of characteristics is demanded.

In each data area,
The overhead as training bits used in the digital PLL circuit is used. However, if the overhead increases, the data area in one burst frame is compressed. Therefore, this overhead is minimized and the data area is efficiently used. It is required to realize high-speed pull-in while using well.

Therefore, for example, in response to the input of burst data having a phase variation due to a duty variation, a jitter, a frequency deviation, etc., a conventional technique for outputting a reproduced data signal which has been subjected to identification retiming at high speed and without error is provided. There is a "digital PLL circuit" disclosed in Japanese Patent Application Laid-Open No. 8-237117.

Here, an embodiment of a "digital PLL circuit" disclosed in the above-mentioned Japanese Patent Application Laid-Open No. 8-237117 will be described with reference to FIG. FIG. 17 is a block diagram showing a configuration of one embodiment of a "digital PLL circuit" disclosed in the above-mentioned Japanese Patent Application Laid-Open No. 8-237117.

As shown in FIG. 17, this digital PLL circuit includes an input terminal 100 to which an input data signal 101 is input, a data sampling circuit 123 for performing sampling of the input data signal 101, and N sampling circuits. A change point of the input data signal 101 (hereinafter, this change point is also referred to as an edge) is detected from the values of the input data signals D0 to DN, and information 10 of the change point is obtained.
7, an edge detection circuit 124 that outputs number information 109 of rising transition points and number information 110 of falling transition points, a falling edge counter 125 that calculates an average value 104 of detected edges, and an extraction clock. A data identification retiming circuit 128 for outputting a reproduced data signal 112 synchronized with the signal 105;
2 as the clock synchronized with 2
And a clock selector 127 that outputs the clock signal 05.

This digital PLL circuit samples input data with an N-phase clock signal composed of N clock signals whose phases are shifted at equal intervals, and extracts N sample data signals D0 to DN. Further, the edge of the input data signal is detected by further sampling the extracted N sample data signals at a predetermined clock.

Here, the operation of the conventional digital PLL circuit shown in FIG. 17 will be described with reference to FIG. 17 and FIG.

FIG. 14 shows the digital P shown in FIG.
FIG. 3 shows a conceptual diagram of the operation of the LL circuit. However, in the example shown in FIG. 14, the number of phases of the N-phase clock signal is 8 (accordingly, N
= 8) will be described as an example. The edge detection operation described with reference to FIG. 14 is also used in the present invention.

When the input data signal is sampled at a certain time by an eight-phase clock signal, the sampled data is composed of "0" and "1" like the sample data shown in FIG. It becomes a continuous data string.

Therefore, the sample data is changed from 0 to 1
Is changed to a rising change point (edge),
A case where the value changes from 1 to 0 is defined as a falling change point (edge).

Then, in order to number these rising transition points and falling transition points, in the case of a rising transition from 0 to 1, a clock indicating 1 data immediately after rising to 1 is output. A phase number (“2” in the case of FIG. 14) is assigned, and in the case of a falling transition from 1 to 0, a phase number of a clock indicating data of 0 immediately after the falling to 0 (FIG. 14) In the case of, "7") is added.

The edge detection circuit 124 detects the edge of the input data signal by the above-described operation. Furthermore, along with the detection of the edge, in one signal cycle of the extracted clock,
Information on the number of rising transition points and the number of falling transition points is also calculated.

The edge detection circuit 124 shown in FIG. 17 converts the clock phase number information 107 shown in FIG.
7, and outputs information 109 and 110 on the number of rising transition points and the number of falling transition points to the data identification retiming circuit 128.

The falling edge counter 125 is a circuit for calculating the average value of the falling clock phase numbers.

If there is no error in the input data signal, that is, if there is no phase fluctuation due to duty distortion or jitter, it is not necessary for the falling edge counter 125 to take an average value. .

However, in reality, since the input data signal has a phase fluctuation due to jitter or the like, the input data signal has a phase difference of
The phase number of a rising clock that changes from 1 to 1 and the phase number of a falling clock that changes from 1 to 0 vary.

Therefore, the falling edge counter 125
, The average value of the phase numbers of falling clocks from the past to the present is taken. In addition, as shown in FIG. 17, such an average value will not generally be an integer, so by rounding this average value,
Adjust to an integer value.

The operation of calculating the average value in the falling edge counter 125 is based on the phase (phase information) of the center value of the fluctuation (jitter) in the time axis direction at the changing point of the input data signal.
Is obtained for the falling transition point.

Therefore, when the center value of the fluctuation (jitter) fluctuates at a low speed, the average value follows the operation.

Finding the average value means suppressing fluctuations (jitter) at the changing points of the input data signal. In the characteristics of the digital PLL circuit, the role of the average value is to perform an operation of suppressing high-frequency components of fluctuation (jitter) and following low-frequency components.

Then, the falling edge counter 125
Outputs the average value 104 representing clock phase information to the clock selector 127. Clock selector 127
Selects a clock signal corresponding to this average value, and uses the data sampling circuit 12 as an extracted clock signal.
3. Output to the data identification retiming circuit 128 and the outside.

On the other hand, the extracted clock input to the data sampling circuit 123 plays a role of a predetermined clock when further sampling the N sample data signals.

The data identification retiming circuit 128 includes:
Information 109 on the number of rising transition points output from the edge detection circuit 124, information 1 on the number of falling transition points
10. N output from the data sampling circuit 123
The book sample data signal and the extracted clock signal output from the clock selector 127 are input.

Here, the data reproducing operation in the data identification retiming circuit 128 shown in FIG.
This will be described with reference to FIG.

FIG. 15 shows a conceptual diagram of the data reproduction operation in the data identification retiming circuit 128 shown in FIG. 17, but the operation concept described below is also used in the present invention.

As shown in FIG. 15, the data reproducing method in the data identification retiming circuit 128 is controlled by the number of the change points of the input data signal during one cycle of the extracted clock signal.

That is, as is apparent from FIG.
The data discrimination retiming circuit 128 includes the number information 1 of rising transition points and falling transition points in one cycle of the extracted clock, which is output from the edge detection circuit 124.
09 and 110 are input, so using this information,
Determine the value of the playback data signal.

For example, when the number of change points is 0, the received data during that time is always a steady value of 0 or 1 (pattern (A) in FIG. 15).

When the number of change points is two, the received data between them has one upwardly convex pulse or one downwardly convex pulse (see the pattern (FIG. 15)). B)).

Here, when the number of change points is one, the received data during that change always changes during the clock cycle. Then, in this case, when the position of the falling transition point is less than 0.5T, it is identified as 0, and when the rising transition point is at a position of 0.5T or more, it is identified as 0 (see FIG. 15). Pattern (C)).

By the above operation, the data identification retiming circuit 128 shown in FIG. 17 can output the reproduced data signal 112 synchronized with the extracted clock signal.

Therefore, in the conventional digital PLL circuit and signal reproducing method, even if the phase of the input data fluctuates due to jitter or the like, the operation of calculating the average value of the change points of the input data is performed. Data is sampled in the time direction by the multi-phase clock signal,
Since retiming is performed after data identification based on the number of data edges, it is possible to synchronize with the input data signal in a short time of several bits with respect to the input of burst data having phase variation due to frequency deviation, duty variation, jitter, etc. Extract the clock signal as the extracted clock signal,
It is stated that a digital PLL circuit and a signal reproducing method capable of outputting data subjected to identification retiming without error as a reproduced data signal can be provided.

[0051]

SUMMARY OF THE INVENTION Conventional digital PL
In the L circuit and the signal reproducing method, the speed-up of the pull-in time is still not enough, and furthermore, a high-speed pull-in is realized, and the data area is used as efficiently as possible. Digital PLL with sufficient resistance to noise and duty distortion
It is required that the circuit and the signal reproducing method be constructed economically and inexpensively.

However, in the conventional digital PLL circuit and signal reproducing method, the high-speed pull-in and the efficient use of the data area, and the high-speed pull-in and the tolerance against jitter fluctuation and duty distortion are required. Digital PLL circuits and signals that use data areas as efficiently as possible while realizing high-speed pull-in and have sufficient tolerance against jitter fluctuation and duty distortion of data signals input to the digital PLL circuit. There is a problem that a reproduction method cannot be provided.

Here, in the following description, in the conventional digital PLL circuit and signal reproducing method, high-speed pull-in and efficient use of the data area, and high-speed pull-in and tolerance to jitter fluctuation and duty distortion. Will be described.

<Relationship between High-speed Lock-in and Efficient Use of Data Area> In the conventional digital PLL circuit and signal reproducing method, in order to shorten the lock-in time (thus, reduce the number of bits). Figure 1
As shown in FIG. 3, it is conceivable to increase the number of overhead bits used for clock signal reproduction and the like.

By increasing the number of bits of the overhead, the first number of bits at which there is no identification error in the reproduced data signal decreases, and high-speed pull-in can be realized.

However, when the number of bits of the overhead is increased, an area usable as a data area is compressed. Therefore, the high-speed pull-in and the efficient use of the data area are in conflict with each other and cannot be realized at the same time.

<Relationship between high-speed pull-in and tolerance to jitter fluctuation and duty distortion> In a conventional digital PLL circuit, generally, in order to increase jitter fluctuation tolerance and duty fluctuation tolerance, phase fluctuation of an input data signal is required. On the other hand, it is necessary to suppress the amount of phase correction and perform feedback control with a small gain.However, in such a feedback circuit having a small gain with respect to the input phase, the amount of phase correction is small per unit time of phase comparison. The retraction time increases.

On the other hand, in order to shorten the pull-in time,
It is necessary to perform feedback control with a large gain for the phase fluctuation of the input data signal.However, in such a feedback circuit with a large gain for the input phase, the phase information drawn in largely follows jitter fluctuation and duty fluctuation. Alternatively, oscillation occurs, causing an error in data identification, and as a result, the profit of the jitter fluctuation body and the tolerance for duty fluctuation are reduced.

Therefore, in the conventional digital PLL circuit, high-speed pull-in and the ability to withstand jitter fluctuation and duty distortion are contradictory operations. If the pull-in time is further increased, the jitter fluctuation and duty cycle of the input data signal are reduced. There is a problem in that the tolerance to distortion is reduced, and the reproduced data signal is likely to be erroneous.

The present invention has been made in view of the above circumstances, and it is an object of the present invention to efficiently use a data area and to speed up a pull-in time without lowering a tolerance against jitter fluctuation and duty distortion of an input data signal. It is an object of the present invention to provide a digital PLL circuit and a signal reproducing method that can perform the above.

[0061]

According to the first aspect of the present invention,
In a digital PLL circuit that outputs a reproduced data signal synchronized with the extracted clock signal, the reproduced data signal synchronized with the extracted clock signal is output at high speed and with a small error by delaying the sample data signal.

According to a second aspect of the present invention, the sampled data signal obtained by sampling the input data signal with the N-phase clock signal is delayed, and the time until the extracted clock signal is selected is reduced. The reproduction data signal synchronized with the extracted clock signal is output at high speed and with a small error by setting the time to reproduction or less.

According to a third aspect of the present invention, at least a data sampling processing section for sampling an input data signal and outputting N sample data signals, and a data reproducing section for outputting a reproduced data signal synchronized with an extracted clock signal. A digital PLL circuit having a processing unit, a delay processing unit for delaying the N sample data signals between the data sampling processing unit and the data reproduction processing unit, wherein the extracted clock signal is selected. By setting the time until the reproduction data signal is equal to or less than the time until the reproduction data signal is reproduced, the reproduction data signal synchronized with the extracted clock signal is output at high speed and with a small error.

According to a fourth aspect of the present invention, an input data signal and an N-phase clock signal composed of N clock signals are input, and the input data signal is sampled by the N-phase clock signal. A data sampling processing unit that outputs N sample data signals; and a data reproduction processing unit that outputs a reproduced data signal synchronized with the extracted clock signal, wherein the reproduced data signal is selected from the N-phase clock signals. In the digital PLL circuit that outputs an extracted clock signal synchronized with the N sample data signal, the N sample data signals output from the data sampling processing unit are input, and the input N sample data signals are delayed. As a delayed sample data signal of
A delay processing unit that outputs the input data signal to the data reproduction processing unit, and the input data signal is reproduced as a reproduction data signal during a time until the extracted clock signal is selected and output based on the input data signal. By setting the time to be less than or equal to the time, a reproduced data signal synchronized with the extracted clock signal is output at high speed and with a small error.

According to a fifth aspect of the present invention, in the fourth aspect, the N-phase clock signal has substantially the same frequency as the input data signal, and has a phase of 360 degrees / N (N is 2 or more). (Integer), and is a signal composed of N clock signals sequentially shifted by an integer.

According to a sixth aspect of the present invention, in the fourth or fifth aspect of the present invention, the data reproduction processing section selects one of the N delayed sampled data signals output from the delay processing section. Select the delayed sample data signal and
It is characterized in that it is output as a reproduced data signal in synchronization with the extracted clock signal.

According to a seventh aspect of the present invention, in the invention according to any one of the fourth to sixth aspects, an extracted clock selection processing unit for selecting and outputting the extracted clock signal from the N-phase clock signals is provided. It is characterized by having.

According to an eighth aspect of the present invention, in the invention according to any one of the fourth to seventh aspects, the delay processing section comprises N
M stages (M
Is an arbitrary integer of 1 or more), and has N flip-flop lines, and a clock signal input to a flip-flop constituting each flip-flop line is a clock constituting the N-phase clock signal. It is a signal.

According to a ninth aspect of the present invention, in the invention according to any one of the fourth to seventh aspects, the delay processing section comprises N
M stages (M
Is an integer of 1 or more), N flip-flop lines, and N
An L frequency dividing circuit that divides each of the phase clock signals by 1 / L (L is an arbitrary integer of 2 or more), and the clock signal input to the flip-flop constituting each flip-flop line is The clock signal constituting the N-phase clock signal is a signal obtained by dividing the frequency by the L frequency dividing circuit.

The tenth aspect of the present invention provides the fourth to ninth aspects.
In the invention according to any one of the above, the N sample data signals output from the data sampling processing unit and the extracted clock signal are input, and the N sample data signals and the extracted clock signal are input to the input N sample data signals and the extracted clock signal. Based on
The image processing apparatus further includes a change point detection calculation section that outputs a signal for controlling the extraction clock selection processing section and the data reproduction processing section as a change point calculation output signal.

According to an eleventh aspect of the present invention, in accordance with the tenth aspect of the present invention, the change point operation output signal indicates a falling change point of the input data signal. It is characterized by including phase number information.

A twelfth aspect of the present invention is the clock signal forming the N-phase clock signal according to the tenth or eleventh aspect, wherein the change point operation output signal indicates a rising change point of the input data signal. It is characterized by including information on the phase number of the signal.

According to a thirteenth aspect of the present invention, in any one of the tenth to twelfth aspects, the change point calculation output signal includes information indicating the number of change points of the input data signal. And

According to a fourteenth aspect of the present invention, in the invention according to any one of the tenth to thirteenth aspects, the change point detecting and calculating section converts the N sample data signals into a timing synchronized with the extracted clock signal. Calculating a phase number of a clock signal corresponding to a rising transition point and a falling transition point of the input data signal from the received N values, and calculating an average value of the detected phase numbers. Is performed to output the change point calculation output signal.

According to a fifteenth aspect of the present invention, in the invention according to any one of the tenth to fourteenth aspects, the extracted clock selection processing unit outputs the change point calculation output signal output from the change point detection calculation unit. The clock of the phase indicated by the average value of the phase number of the clock signal indicating the rising transition point of the input data signal or the average value of the phase number of the clock signal indicating the falling transition point of the input data signal By selecting a signal from the N-phase clock signal and outputting it as an extracted clock signal, it is possible to output an extracted clock signal that follows a change in a rising transition point or a falling transition point of an input data signal. Features.

The invention according to claim 16 is the invention according to claims 4 to 1
5. The invention according to any one of 5, wherein the delay processing section delays the input N sample data signals while maintaining a phase difference between the N sample data signals.

According to a seventeenth aspect of the present invention, the input data signal has substantially the same frequency as the input data signal, and has a phase of three.
An N-phase clock signal composed of N clock signals sequentially shifted by 60 degrees / N (N is an integer of 2 or more) is input, and the input data signal is sampled by the N-phase clock signal. A data sampling processor (1) for outputting N sample data signals, N sample data signals (6) output from the data sampling processor, and the N-phase clock signal; The N sampled data signals are delayed based on the N-phase clock signal, and output as N delayed sampled data signals (7). The sample data signal of the book and the extracted clock signal output from the extracted clock selection processing unit (5) are input, and the extracted clock selection processing unit; A change point detection operation section (4) for outputting a change point operation output signal (8) for controlling the data reproduction processing section (3); A clock signal to be output as an extracted clock signal is selected from the N-phase clock signal based on the output change point calculation output signal, and the selected clock signal is used as the extracted clock signal. An extracted clock selection processing unit (5) to output, the N delayed sample data signals output from the delay processing unit, a change point calculation output signal output from the change point detection calculation unit, and the extracted clock A data reproduction processing unit (3) that receives the extracted clock signal output from the selection processing unit and outputs a reproduction data signal.

According to the eighteenth aspect of the present invention, in the signal reproducing method for outputting a reproduced data signal synchronized with the extracted clock signal, the sample data signal is delayed by
The reproduction data signal synchronized with the extraction clock is output at high speed with a small error.

According to a nineteenth aspect of the present invention, the sampled data signal obtained by sampling the input data signal by the N-phase clock signal is delayed, and the time until the extracted clock signal is selected is reduced. The reproduction data signal synchronized with the extracted clock signal is output at high speed and with a small error by setting the time to reproduction or less.

According to a twentieth aspect of the present invention, at least a data sampling process of sampling an input data signal and outputting N sample data signals, and a data reproduction process of outputting a reproduced data signal synchronized with an extracted clock signal A signal reproducing method having a processing step, wherein a delay processing step of delaying the N sample data signals is provided between the data sampling processing step and the data reproduction processing step, and the extracted clock signal is selected. By setting the time until the reproduction data signal is equal to or less than the time until the reproduction data signal is reproduced, the reproduction data signal synchronized with the extracted clock signal is output at high speed and with a small error.

According to a twenty-first aspect of the present invention, an input data signal and an N-phase clock signal composed of N clock signals are input, and the input data signal is sampled by the N-phase clock signal. A data sampling processing step of outputting N sample data signals; and a data reproduction processing step of outputting a reproduction data signal synchronized with the extracted clock signal, wherein the reproduction data signal is selected from the N-phase clock signals. In the signal reproducing method for outputting an extracted clock signal synchronized with the above, the N sample data signals output from the data sampling processing step are input, the input N sample data signals are delayed, A delay processing step of outputting as a delayed sampled data signal of which the extracted clock signal is based on the input data signal. The time until the output is-option,
By setting the input data signal to be equal to or shorter than the time required until the input data signal is selected as a reproduced data signal, a reproduced data signal synchronized with the extracted clock signal is output at high speed and with a small error.

According to a twenty-second aspect, in the twenty-first aspect, the N-phase clock signal has substantially the same frequency as the input data signal, and has a phase of 360 degrees / N.
(N is an integer of 2 or more) N-phase clock signals sequentially shifted by N clock signals.

According to a twenty-third aspect of the present invention, in the invention of the twenty-first or twenty-second aspect, the data reproduction processing step includes selecting one of the N delayed sampled data signals output in the delay processing step. The delay sample data signal is selected and output as a reproduced data signal in synchronization with the extracted clock signal.

According to a twenty-fourth aspect of the present invention, in the invention according to any one of the twenty-first to twenty-third aspects, there is provided an extracted clock selecting step of selecting and outputting the extracted clock signal from the N-phase clock signals. It is characterized by having.

According to a twenty-fifth aspect of the present invention, in the first aspect of the present invention, the delay processing step includes a step of converting each of the N sample data signals into M stages (M is an arbitrary integer of 1 or more). ), The clock signal constituting the N-phase clock signal is input to the N flip-flop lines constituted by the flip-flops, and the clock signal constituting the N-phase clock signal is used as the clock signal inputted to the flip-flop constituting each flip-flop line. This is a step of delaying the input N sample data signals.

According to a twenty-sixth aspect of the present invention, in the first aspect of the present invention, the delay processing step includes a step of performing N stages of sample data signals in each of M stages (M is an arbitrary integer of 1 or more). ) Is inputted to the N flip-flop lines constituted by the flip-flops, and each of the inputted N-phase clock signals is divided by 1 / L (L is an arbitrary integer of 2 or more). The clock signal constituting the N-phase clock signal is divided by the L divider circuit as a clock signal to be inputted and inputted to the flip-flop constituting each flip-flop line. This is a step of delaying the sample data signal of the book.

According to a twenty-seventh aspect of the present invention, in the first aspect of the present invention, based on the N sample data signals output in the data sampling processing step and the extracted clock signal, And a change point detection calculation step of outputting a signal for controlling the extraction clock selection processing step and the data reproduction processing step as a change point calculation output signal.

According to a twenty-eighth aspect of the present invention, in the twenty-seventh aspect of the present invention, the change point operation output signal indicates a falling change point of the input data signal. It is characterized by including phase number information.

The invention according to claim 29 is the clock according to claim 27 or 28, wherein the change point calculation output signal indicates a rising change point of the input data signal and constitutes the N-phase clock signal. It is characterized by including information on the phase number of the signal.

According to a thirtieth aspect of the present invention, in any one of the twenty-seventh to twenty-ninth aspects, the change point calculation output signal includes information indicating the number of change points of the input data signal. And

According to a thirty-first aspect of the present invention, in the invention according to any one of the twenty-seventh to thirty-seventh aspects, the change point detecting operation step includes a step of synchronizing the N sample data signals with the extracted clock signal. Calculating a phase number of a clock signal corresponding to a rising transition point and a falling transition point of the input data signal from the received N values, and calculating an average value of the detected phase numbers. Is performed to output the change point calculation output signal.

The invention according to claim 32 is the invention according to any one of claims 27 to 31, wherein the extracted clock selection processing step is the change point calculation output signal output in the change point detection calculation step. The clock of the phase indicated by the average value of the phase number of the clock signal indicating the rising transition point of the input data signal or the average value of the phase number of the clock signal indicating the falling transition point of the input data signal By selecting a signal from the N-phase clock signal and outputting it as an extracted clock signal, it is possible to output an extracted clock signal that follows a change in a rising transition point or a falling transition point of an input data signal. Features.

According to a thirty-third aspect of the present invention, in the invention according to any one of the twenty-first to thirty-second aspects, the delay processing step comprises the step of receiving the input data while maintaining a phase difference between the N sample data signals. The method is characterized in that N sample data signals are delayed.

[0094]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, a digital P according to the present invention will be described.
An embodiment of an LL circuit and a signal reproducing method will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing a digital PLL circuit according to an embodiment of the present invention. However, among the members and signals shown in FIG. 1, the members provided in the conventional digital PLL circuit shown in FIG.

As shown in FIG. 1, this digital P
The LL circuit includes a data sampling processing unit 1, a delay processing unit 2, a data reproduction processing unit 3, a change point detection calculation unit 4, and an extracted clock selection processing unit 5.

The data sampling section 1 receives an input data signal and an N-phase clock signal, and digitally samples the input data signal using the N-phase clock signal. The signal 6 is output.

The delay processing section 2 receives the sample data signal 6 output from the data sampling processing section 1 and the N-phase clock signal, and delays the sample data signal 6 composed of N sample data signals. , And N delayed sampled data signals.

Here, the N-phase clock signal has substantially the same frequency as the input data signal, and has a phase of 360 degrees / N (N is 2
It is composed of N clock signals sequentially shifted by the above integers).

In the following description, among the N-phase clock signals, a clock signal whose phase is not shifted with respect to the input data signal is defined as a zero-phase clock signal, and the phase of the input data signal is (360 / N) × The clock signal shifted by 1 is used as the first phase clock signal, and the phase of the input data signal is (3
A clock signal shifted by 60 / N) × 2 is referred to as a second-phase clock signal, and the clock signal shifted in phase by (360 / N) × n from the input data signal is referred to as an n-th phase clock signal. I do.

The change point detection / calculation section 4 receives the sample data signal 6 output from the data sampling processing section 1 and the extracted clock signal output from the extracted clock selection processing section 5, and inputs a change point calculation output signal 8 Is output.

The extracted clock selection processing unit 5 receives the N-phase clock signal and the change point calculation output signal 8 output from the change point detection calculation unit 4, and outputs an extracted clock signal from the N-phase clock signals. I do.

The data reproduction processing unit 3 includes a delay sample data signal 7 output from the delay processing unit 2, a change point operation output signal 8 output from the change point detection operation unit 4, and an extraction clock selection operation unit 5. The output extracted clock signal is input, and a reproduced data signal is output.

Each member shown in FIG. 1 will be described in more detail below. First, the data sampling processing unit 1
Samples the input data signal digitally with an N-phase clock signal, and converts a sample data signal 6 composed of N sample data signals into a delay processing unit 2.
And the change point detection calculation unit 4.

Here, of the sample data signal 6 composed of the N sample data signals, the sample data signal sampled and output by the clock signal of the 0th phase is converted to the sample data signal of the 0th phase. The sample data signal sampled and output by the one-phase clock signal is referred to as a first-phase sample data signal, and similarly, the sample data signal sampled and output by the n-th clock signal is referred to as an n-th sample signal. Name it a data signal.

The change point detection calculation section 4 takes in the sample data signal 6 at a timing synchronized with the extracted clock signal, for example, at a rising timing of the extracted clock signal. Here, the extracted clock signal is selected and output from the extracted clock selection processing unit 5 based on the change point calculation output signal 8, but in the initial operation where the change point calculation output signal 8 is not output, the N-phase Any one clock signal is selected from the clock signals.

The above-described acquisition of the sample data signal 6 at the timing synchronized with the extracted clock signal means that the sample data signal of the 0th phase to the sample data of the Nth phase is obtained at the rising time of the extracted clock signal. This means taking in the data.

Then, the acquired sample data signal 6
The phase information indicating the positions of the rising transition point and the falling transition point of the input data signal is included in the change point calculation output signal 8 and output.

The capture of the sample data signal 6 in the change point detection calculation unit 4 has been partially described with reference to FIG. 14 in the above-described conventional technique, but will be described in more detail below with reference to FIG. Will be described. However, as shown in FIG. 2, the data indicated by the input data signal is
.., -1, 0, 1, 2,... Are sequentially numbered in chronological order. In the following, data is simply identified by these numbers.

First, it is assumed that the first-phase clock signal C1 is selected as the extraction clock to be input to the change point detection calculation section 4. In this case, as shown by the line A in FIG. 2, the value of each sample data signal is taken in at the rise of the first phase clock signal C1.

When the value of each sample data signal is taken in at the rise of the first phase clock signal C1, the values from the 0th phase sample data signal D0 to the 7th phase sample data signal D7 are sequentially 0,-. 1, -1, 0, 0,
0, 0, 0, and the time when the input data signal rises from -1 to 0 is the time when the third phase clock signal C3 rises. When the input data signal is 0
From the second phase clock signal C
It can be seen that this is the time point of the rise of No. 2.

Also, as shown in line B of FIG.
When the value of each sample data signal is taken in at the fall of the first phase clock signal C1, the zero-phase sample data signal D0 is converted to the seventh phase sample data signal D7.
Are sequentially 0, 0, 0, 1, 1, 0, 0, 0, and it can be seen that the time when the input data signal rises from 0 to 1 is the time when the third phase clock signal C3 rises. . Then, it can be seen that the time when the input data signal falls from 1 to 0 is the time when the second phase clock signal C2 rises.

Next, it is assumed that the clock signal C4 of the fourth phase is selected and input as the extracted clock signal to be input to the change point detection calculation section 4. In this case, as shown by the line C in FIG. 2, when the value of each sample data signal is taken in at the rise of the clock signal C4 of the fourth phase, the sample data signal D0 of the zero phase becomes The values up to the sample data signal D7 are sequentially 1, 1, 1,
2, 1, 1, 1, 1 and the input data signal is 1 to 2
At the rising edge of the third phase clock signal C3. Then, it can be seen that the time when the input data signal falls from 2 to 1 is the time when the second phase clock signal C2 rises.

Also, as shown in line D of FIG.
When the value of each sample data signal is taken in at the falling of the fourth phase clock signal C4, the seventh phase sample data signal D7 is converted from the zero phase sample data signal D0.
Are sequentially 1, 1, 1, 2, 2, 2, 2, 2. It can be seen that the time when the input data signal rises from 1 to 2 is the time when the third phase clock signal C3 rises. . Then, it can be seen that the time when the input data signal falls from 2 to 1 is the time when the second phase clock signal C2 rises.

As can be seen from the above-described example, the change point detection calculation unit 4 captures the value of each sample data signal in synchronization with the input extracted clock signal, and calculates the rising edge of the input data signal from the captured value. , And the falling transition point are detected as the phase numbers of the clock signals constituting the N-phase clock signal. Further, information on the number of rising transition points and the number of falling transition points is calculated.

Next, the change point detection calculation section 4 performs a process of obtaining an average value for each of the phase number indicating the rising change point and the phase number indicating the falling change point.

The process of obtaining the average value includes the process of obtaining the average value of the phase number indicating the rising transition point from the past to the current value, and the process of obtaining the average value of the phase number indicating the falling transition point from the past to the current value. The average value of the phase number indicating the rising transition point and the average value of the phase number indicating the falling transition point are calculated by performing the two processes of taking the value.

The operation of calculating the average value will be described with reference to FIG. FIG. 3 shows a conceptual diagram of the operation of the averaging process performed by the change point detection calculation unit 4. Here, the processing for obtaining the average value of the phase numbers indicating the rising transition points will be described. However, the processing for obtaining the average values of the phase numbers indicating the falling transition points can be performed in exactly the same manner as described below.

First, the phase number indicating the rising change point of the input data signal detected by the change point detection calculation section 4 is input to the subtractor 201, and the averaging section (ACC) 204
Is subtracted from the already averaged value output from.

Thereafter, X is set to the value of the phase number after the subtraction, and the weighting section 202 performs weighting by the function f (X). The weighting by the function f (X) may be linear weighting such as f (X) = (1/4) X, any other weighting, or weighting according to the elapsed time. good.

The output after weighting is added to the adder 20.
3 is added to the already averaged value output from the averaging unit 204 in the adder 203, and then output to the averaging unit (ACC) 204.

After performing the normal averaging process, the averaging unit 204 outputs the average value of the phase number indicating the rising change point.

Here, the value after the averaging process is generally not an integer, and as it is, it is not practical as a clock phase number indicating a rising transition point or the like. Therefore, the value after this averaging process is rounded,
Make it an integer. By doing so, it becomes practical to use as a phase number of a clock indicating a position such as a rising transition point.

The change point detecting and calculating section 4 calculates the value indicating the average value of the phase numbers of the rising change points and the value indicating the average value of the phase numbers of the falling change points, obtained as described above.
And the number of rising transition points and the number of falling transition points, as shown in FIG.
To the extracted clock selection processing unit 5 and the data reproduction processing unit 3.

Next, the extracted clock selection processing section 5 outputs a change point calculation output signal 8 output from the change point detection calculation section 4.
A value indicating the average value of the phase numbers of the rising transition points,
And a value indicating the average value of the phase number of the falling transition point, and selecting the signal of the phase indicated by the value from the N-phase clock signal based on one of the values. Is output as an extracted clock signal that follows the fluctuation of

Next, the operation of the delay processing section 2 will be described. The delay processing unit 2 inputs the sample data signal 6,
Each of the N signals constituting the sample data signal 6 is delayed using an N-phase clock signal while maintaining its phase relationship, and is output as a delayed sample data signal 7.

Here, the configuration of the delay processing section 2 shown in FIG. 1 will be described with reference to FIG. FIG. 4 shows a block diagram of the configuration of the delay processing unit 2 shown in FIG.

As shown in FIG. 4, the delay processing unit 2
The sample data signal 6 is branched into N sample data signals D0 to DN-1, which are M-stage (M is an arbitrary integer of 1 or more) flip-flop circuits 11-1 using an N-phase clock signal. -1 to 11-NM, and outputs N delayed sample data R0 to RN-1. Thereafter, the N delayed sampled data signals are combined and output as a delayed sampled data signal 7.

Here, each sample data signal D
A line composed of M stages of flip-flops to which 0 to DN-1 is input is called a flip-flop line.

Next, the data reproduction processing unit 3 shown in FIG.
Will be described. The data reproduction processing unit 3 determines the value of the reproduction data signal using the change point calculation output signal 8, the delayed sample data signal 7 output from the delay processing unit 2, and the extracted clock signal, and The reproduction data signal is output at a timing synchronized with the reproduction data signal.

In the prior art, as described with reference to FIG. 15, the reproduction data processing unit 3 controls the reproduction data signal to be output based on the number of transition points.

Next, the operation of the digital PLL circuit shown in FIG. 1 and the signal reproducing method using the same will be described in more detail with reference to the drawings.

First, the operation of the data sampling processing section 1 will be described with reference to FIG. FIG. 6 shows a time chart of the operation of the data sampling processing section 1 shown in FIG. However, in the following description, a case where the number of phases N of the N-phase clock signal is eight will be described.

As shown in FIG. 6, the input data signal input to the data sampling processor 1 is sampled at the rising transition point of the 0th phase clock signal C0 to the 7th phase clock signal C7, The phase sample data signal D0 is output as the seventh phase sample data signal D7.

Each of the 0th phase clock signal C0 to the 7th phase clock signal C7, as shown in FIG.
There is a phase relationship that is sequentially shifted at intervals obtained by equally dividing one cycle T of the clock into eight.

Next, the operation of the delay processing section 2 will be described with reference to FIG.
This will be described with reference to FIG. FIG. 7 shows a time chart of the operation of the delay processing unit 2 shown in FIG. However, the case shown in FIG. 7 will be described below with an example where the number of delay stages M is four.

Eight phase sample data signals D0 to D0 sampled by each of the eight phase clock signals
D7 is delayed by four stages in the delay processing unit 2 in the flip-flop lines while maintaining the original sample data signals D0 to D7 in accordance with the eight-phase clock signal, thereby delaying the 0th phase delay sample. Data signal R0
Is output as the seventh-phase delayed sample data signal R7.

FIG. 7 shows a sample data signal D of the third phase.
3 is delayed by four stages to become a third-phase delayed sample data signal R3, with the data “0” being emphasized.

Next, the change point detecting and calculating section 4 shown in FIG.
Will be described with reference to FIG. FIG. 8 shows a time chart of the operation of the change point detection calculation section 4 shown in FIG.

The change point detection / calculation section 4 takes in the sample data signal D0 of the 0th phase to the sample data signal D7 of the 7th phase in synchronization with the extracted clock signal, and calculates the rising change of the input data signal from the taken value. Point and the phase number of the clock indicating the falling transition point are detected,
In addition to calculating the average value of the phase numbers from the past to the present, the number of rising transition points and the number of falling transition points during one cycle of the extraction clock were calculated and synchronized with the extraction clock signal. An operation of outputting the change point calculation output signal 8 as a signal representing those values at the timing is performed.

Normally, processing time is required from the detection of the value indicating the phase number of the rising transition point and the value indicating the phase number of the falling transition point to the calculation for calculating the average value. Here, the case where the processing time ends in one cycle of the extracted clock signal is shown.

From the input data signal, the change point calculation output signal 8
The delay time until is as shown in FIG. However, in FIG. 8, for example, “data up to −1”
When the input data signal is -1, the clock phase number indicating the average value of the rising transition points and the average value of the falling transition points, the number of rising transition points and the falling transition during one cycle of the extracted clock. Indicates the number of points.

Here, the expression “until” means that
This is because the average value of the rising change point and the average value of the falling change point of the input data signal are determined based on the respective values of the past input data signal. Also, in FIG. 8, the manner in which “data up to 0” is output with a delay is emphasized.

Next, the operation of the extracted clock selection processing section 5 shown in FIG. 1 will be described with reference to FIG. FIG.
2 shows a time chart of the operation of the extracted clock selection processing unit 5 shown in FIG.

The extracted clock selection processing section 5 has a value indicating the phase number of the rising change point indicated by the change point calculation output signal 8,
According to the information of one of the phase numbers of the falling transition points and the values indicating the phase numbers of the falling transition points, a clock signal of one phase is selected as an extracted clock signal from among the eight-phase clock signals and output as needed.

In FIG. 9, the average value of the phase numbers of the clock signal indicating the falling transition point is used as the information used by the extracted clock selection processing unit 5. Then, in the “data up to −2” of the change point calculation output signal 8, the average value of the falling transition points indicates three phases, and in the “data up to −1”, the average value of the falling transition points is The phase information indicated by the change point calculation output signal 8 sequentially includes two phases,
This shows how the extracted clock signal changes when the phase changes to three or four.

Next, the data reproduction processing unit 3 shown in FIG.
Will be described with reference to FIG. FIG.
2 shows a time chart of the operation of the data reproduction processing unit 3 shown in FIG.

The signal input to the data reproduction processing unit 3 is
While the phase relationship between the sample data signals is maintained by the delay processing unit 2, each of the flip-flop lines delays by four stages from the 0th-phase delayed sample data signal R0 to the 7th.
The value indicating the average value of the phase number of the rising transition point, the value indicating the average value of the phase number of the falling transition point, and the value of the rising transition point, which are output from the phase delay sample data signal R7 and the transition point detection calculation unit 4. And the extracted clock signal representing the number of falling transition points and the number of falling transition points.

The data reproduction processing section 3 selects one delayed sampled data signal from the 0th phase delayed sampled data signal R0 to the 7th phase delayed sampled data signal R7 in accordance with the transition point calculation output signal 8, An operation of outputting this as a reproduced data signal in synchronization with the extracted clock signal is performed.

Since the zero-phase delayed sampled data signal R0 to the seventh-phase delayed sampled data signal R7 have a longer delay time by three stages than the change point operation output signal 8, FIG. As shown, when reproducing the data of “0” of the input data signal, the change point calculation output signal 8 calculated using “data up to +2” of the input data signal and “+2” of the input data signal And the extracted clock signal selected using "data up to".

Here, when there is no data before the data “0” of the input data signal, that is, “−1”, “−”
,... Are all “0” values (the input data signal is a binary value of “0” and “1”), when the first data “0” is reproduced, Already "+
The calculation processing using the data up to "2" is completed.

Next, the data reproduction processing unit 3 shown in FIG.
Will be described in more detail with reference to FIG. 11 and FIG.

FIG. 11 is a block diagram showing an example of the configuration of the data reproduction processing unit 3 shown in FIG. 1, and FIG.
1 shows a conceptual diagram of the operation of the data reproduction processing unit 3 having the configuration shown in FIG.

First, as shown in FIG. 11, the data reproduction processing unit 3 applies the 0-th delay sample data signal R0 to the (N-1) -th phase delay sample data output from the delay processing unit 2. N number of delayed sample data signals 7 up to the signal RN-1, a change point calculation output signal 8 including the number information of change points output from the change point detection calculation unit 4,
The extracted clock signal output from the extracted clock selection processing unit 5 is input.

The data reproduction processing section 3 includes an encoder 802, a selection circuit 801, a selection circuit 804, and a flip-flop 805.

The delay sample data signal 7 from the 0th phase delay sample data signal to the (N−1) th phase delay sample data signal is input to the encoder 802. And
The encoder 802 receives the delayed sampled data signal 7
Among them, the phase number of the delayed sample data signal having the earliest change point (edge) is output.

The selection circuit 801 includes N delayed sampled data signals from the 0th phase delayed sampled data signal R0 to the N-1th phase delayed sampled data signal RN-1 output from the delay processing unit 2. Enter. Then, based on the signal output from the selection circuit 804, the selection circuit 801 selects and outputs one of the N delayed sampled data signals.

The selection circuit 804 outputs the phase number indicating the first edge position of the data output from the encoder 802 and s
(An integer between 0 and N-1 inclusive, where N is the N of the N-phase clock signal
Same as. ), T (an integer of 0 or more and N-1 or less, N is the same as N of the N-phase clock signal), and a change point calculation output signal 8.

The change point calculation output signal 8 is supplied to the selection circuit 804
Input to the selection control terminal. Then, the selection circuit 804 uses the number of falling transition points and the number of rising transition points included in the input transition point calculation output signal 8.

When the number of falling transition points is one, s is selected as the value output from the selection circuit 804. When the number of rising transition points is one, the output from the selection circuit 804 is selected. T is selected as the value to be
If the number of rising transition points and the number of falling transition points are each one, for a total of two, the phase number output from the encoder 802 is output. If both the number of falling transition points and the number of rising transition points are 0, s is selected (or may be t).

In the selection circuit 801, the zero-phase delayed sample data signal R 0 output from the delay processing unit 2 based on the value (integer value from 0 to N−1) indicated by the selection output signal of the selection circuit 804. From the N delayed sampled data signals RN-1 to RN-1 to select and output one data.

The selected output signal from the selection circuit 801 is retimed by the flip-flop circuit 805 using the extracted clock signal as a clock signal and output as a reproduced data signal.

Next, a specific operation of the data reproduction processing section 3 shown in FIG. 11 will be described with reference to FIG. However, in FIG. 11, s = t = 4 is selected,
An example in which sampling is performed using an 8-phase clock signal will be described.

In FIG. 11, only the falling change point of the change point calculation output signal 8 is the change point calculation output signal 8.
If it is determined from the change point number information contained in the data that there is one, the lower column (T in FIG.
S (“4”) is selected and output to the selection circuit 801.

When s = 4 is selected and a signal representing this value is output to the selection circuit 801, a sample data signal 180 degrees away from the average of the falling transition points is output.

That is, the third sample data signal deviated from the zero-phase delayed sample data signal R0 by (360 degrees / 8) × 4 = 180 degrees.
The phase delay sample data signal R3 is selected and output.

Next, when it is determined that only one rising change point in the change point calculation output signal 8 exists from the change point number information included in the change point calculation output signal 8, (FIG. 12) Upper row of c) (T is one cycle of data)
And t (“4”) is selected and the selection circuit 80
1 is output.

When t = 4 is selected and a signal representing this value is output to the selection circuit 801, a sample data signal 180 degrees apart from the average of the rising transition points is output.

That is, the third sample data signal deviated from the zero-phase delayed sample data signal R0 by (360 degrees / 8) × 4 = 180 degrees.
The phase delay sample data signal R3 is selected and output.

It is determined that one falling change point of the change point calculation output signal 8 exists from the change point number information included in the change point calculation output signal 8, and the rising change point is determined by the change point calculation output signal. When it is determined from the change point number information included in No. 8 that there is one (in the case of FIG. 12B), the position of the first change point of the input data signal output from the encoder 802 is determined. The phase number of the indicated clock signal is selected and output to the selection circuit 801.

Upon receiving this output, the selection circuit 801 outputs a delay sample data signal corresponding to the phase number at which the position of the first change point is detected.

If it is determined from the change point number information that there is no change point (in the case of FIG. 12A), or if it is determined that there are three or more change points, s (“4”) ) Is selected (or t), and a signal corresponding to the middle phase (approximately 180 degrees) of the sampled data signal is output.

Thereafter, the selected output signal from the selection circuit 801 is retimed by the flip-flop circuit 805 using the extracted clock signal as a clock signal, and is output as a reproduced data signal synchronized with the extracted clock signal.

In the above example, s and t are both set to 4, and the sampled data signal of the phase corresponding to the center of 1T 180 degrees away from the transition point is selected, and the output which is retimed to be strong against jitter is selected. S = t = 4 as long as the falling and rising jitter patterns are the same. However, depending on the system and circuit configuration of the communication system, both jitter patterns can be obtained. May not be the same, so s and t may be set to 3 and 5 or other values.

As described above, in the digital PLL circuit and the signal reproducing method according to the present invention shown in FIG. 1, by inserting the delay processing unit 2, The process of calculating the average value of the value indicating the phase number of the transition point and the average value of the value indicating the phase number of the falling transition point, and the process of selecting the extracted clock signal using the calculated average value, Apparently,
It is characterized in that it can be performed prior to the reproduction data signal processing in the data reproduction processing section 3.

Therefore, according to the embodiment of the digital PLL circuit, in order to effectively utilize the data area,
Even if the overhead is reduced, the reproduced data signal synchronized with the extracted clock signal can be reproduced at high speed without error.

In addition, it follows the phase (phase information) of the center value of the fluctuation (jitter) of the input data signal at the rising or falling transition point of the input data signal.
The average value of the rising change point or the falling change point is calculated. Therefore, the extracted clock selected according to the average value can be a signal obtained by phase-synchronizing and phase-following the input data signal (this also absorbs the frequency deviation between the input data signal and the N-phase clock signal).

Since the reproduced data signal is a signal whose data value is identified and output at the timing of the extracted clock signal, the reproduced data signal is a signal synchronized with the extracted clock signal. The pull-in time can be shortened without lowering the tolerance against jitter fluctuation and duty distortion of the input data signal.

However, the digital PLL shown in FIG.
The circuit is an example suitable for applying the digital PLL circuit and the signal reproducing method according to the present invention. The digital PLL circuit and the digital PLL circuit applying the signal reproducing method according to the present invention are shown in FIG. The present invention is not limited to such a digital PLL circuit, and various modifications can be made without changing the gist of the present invention.

For example, the configuration of the delay processing unit 2 shown in FIG. 1 has been described with reference to FIG. 4. However, the delay processing unit 2 applied to the present invention has another configuration as shown in FIG. May be used.

FIG. 5 shows the digital PLL shown in FIG.
FIG. 9 is a block diagram of another embodiment of the delay processing unit 2 included in the circuit. However, members and signals similar to those shown in FIG. 4 and signals are denoted by the same reference numerals.

As shown in FIG. 5, first, the delay processing section 2 branches the input sample data signal 6,
The sample data signal D0 of the 0th phase, the sample data signal D1 of the first phase,..., And the sample data signal DN-1 of the (N-1) th phase.

The N-phase clock signal is divided by the L frequency dividing circuit 1
2, the frequency is divided by 1 into L (L is an arbitrary integer of 2 or more), and the divided clock signal is used to generate a 0-phase sample data signal D0 and a first-phase sample data signal D0.
1,..., The (N-1) th phase sample data signal DN-1
Are respectively delayed by flip-flop circuits 11-1-1 to 11-NM of M stages (M is an arbitrary integer of 1 or more),
It is configured to output N delayed sampled data signals from the 0th phase delayed sampled data signal R0 to the (N-1) th phase delayed sampled data signal RN-1.

In the digital PLL circuit using the delay processing unit 2 shown in FIG. 5 and the signal reproducing method using the same, the digital PLL circuit using the delay processing unit 2 shown in FIG. And the same effect as the signal reproducing method using the same can be obtained. Further, by using the configuration of the delay processing unit 2 shown in FIG.
And the power consumption can be reduced with respect to the delay time.

[0185]

As is apparent from the above description, according to the present invention, a reproduced data signal is reproduced and output by inserting a delay processing section between the data sampling processing section and the data reproduction processing section. In the data reproduction processing section, since the average value calculation processing in the change point detection calculation section can be performed in advance, for example, a burst represented by an optical subscriber system communication apparatus or the like can be used. In a digital PLL circuit that requires a pull-in operation of outputting a reproduced data signal and an extracted clock signal at a high speed from the data signal in a few bits, the data area can be efficiently reduced only by changing the number of delay stages of a delay processing unit to be inserted. And shortening the pull-in time without reducing the tolerance to jitter fluctuations and duty distortion of the input data signal. It is possible to provide a digital PLL circuit.

Similarly, by inserting a delay processing step between the data sampling processing step and the data reproduction processing step, in the data reproduction processing step of reproducing and outputting a reproduced data signal, in the change point detection calculation step, Since the average value calculation processing can be performed in advance, it is possible to perform a high-speed reproduction data signal with a few bits from a burst-like data signal represented by an optical subscriber communication device. In a signal reproduction method that requires a pull-in operation of outputting an extracted clock signal, the data area can be used efficiently by simply changing the number of delay stages in a delay processing step to be inserted. It is possible to provide a signal reproducing method capable of shortening the pull-in time without lowering the resistance to distortion.

As a result, it is possible to provide a digital PLL circuit and a signal reproducing method capable of setting the pull-in time to 0 bits and outputting an error-free reproduced data signal from the first bit of the burst data signal. Can be.

In particular, according to the invention of claim 1 and claim 18, since the sample data signal is delayed,
Based on the extracted clock signal, there is a time delay before the reproduced data signal is reproduced and output,
Digital PLL capable of outputting a reproduced data signal synchronized with an extracted clock signal at high speed and with a small error
A circuit and a signal reproducing method can be provided.

According to the second and nineteenth aspects of the present invention, the time until the sampled data signal is delayed and the extracted clock signal is selected and output is reduced until the reproduced data signal is reproduced. By keeping it below the time,
Digital PLL capable of outputting a reproduced data signal synchronized with an extracted clock signal at high speed and with a small error
A circuit and a signal reproducing method can be provided.

According to the third and twentieth aspects of the present invention, the input data signal is sampled into N sample data signals, and the N sample data signals are used to reproduce the data in synchronization with the extracted clock signal. When outputting a data signal, N sample data signals are delayed, and the time until an extracted clock signal is selected and output is set to be equal to or less than the time until a reproduced data signal is reproduced and output. Therefore, it is possible to provide a digital PLL circuit and a signal reproducing method capable of outputting a reproduced data signal synchronized with the extracted clock signal at high speed with a small error.

According to the present invention, the input data signal is sampled by the N-phase clock signal and output as N sample data signals, and based on the output sample data signal. When outputting a reproduced data signal synchronized with the extracted clock signal, the N sample data signals are delayed and output as N delayed sample data signals.
Since the time until the extracted clock signal is selected and output based on the input data signal is shorter than the time until the input data signal is output as the reproduced data signal,
Digital PLL capable of outputting a reproduced data signal synchronized with an extracted clock signal at high speed and with a small error
A circuit and a signal reproducing method can be provided.

According to the fifth and twenty-second aspects of the present invention, the effects of the fourth and twenty-first aspects are obtained, respectively, and the N-phase clock signal is substantially equal in frequency to the input data signal. Since the signals are the same and are composed of N clock signals whose phases are sequentially shifted by 360 degrees / N at an arbitrary integer of 2 or more, the sampling of the input data signal is performed in time series. It is possible,
A digital PLL circuit and a signal reproducing method capable of easily detecting a change point of an input data signal can be provided.

According to the sixth and twenty-third aspects of the present invention, the effects of the fourth and fifth aspects and the twenty-first and twenty-second aspects are obtained, respectively, and the reproduced data signal is composed of N lines. Since one of the delayed sampled data signals is output by synchronizing with the extracted clock signal, the accuracy of the synchronization of the reproduced data signal with the extracted clock signal can be further improved. A digital PLL circuit and a signal reproducing method can be provided.

According to the seventh and twenty-fourth aspects of the present invention, the effects of the fourth to sixth aspects and the twenty-first to twenty-third aspects are obtained, respectively.
Since the extracted clock signal is selected from N-phase clock signals, a digital PLL circuit and a signal reproducing method capable of further improving the accuracy of phase and frequency synchronization of the extracted clock signal with respect to the input data signal are provided. can do.

According to the inventions of claims 8 and 25, the effects of the inventions of claims 4 to 7 and 21 to 24 are obtained, respectively.
Since the delay of the N sample data signals is caused by the flip-flop, it is possible to provide a digital PLL circuit and a signal reproducing method capable of accurately and reliably delaying the sample data signals.

According to the ninth and twenty-sixth aspects of the present invention, the effects of the fourth and seventh aspects and the twenty-first to twenty-fourth aspects can be obtained, respectively, and the N sample data signals can be obtained. Since the delay is due to the flip-flop, the sampled data signal can be accurately and reliably delayed, and a digital PL capable of reducing the circuit scale and the power consumption with respect to the delay time is provided.
An L circuit and a signal reproducing method can be provided.

According to the tenth and twenty-seventh aspects of the present invention, the effects of the fourth to ninth aspects and the twenty-first to twenty-sixth aspects can be obtained, and the selection of the extracted clock signal can be achieved. And a change point calculation output signal as a signal for controlling the reproduction of the reproduced data signal is output, so that the extracted clock signal synchronized with the input data signal can be output more accurately and reliably. In addition, it is possible to provide a digital PLL circuit and a signal reproducing method capable of further accurately synchronizing a reproduced data signal with an extracted clock signal.

According to the eleventh and twenty-eighth aspects of the present invention, the effects of the tenth and twenty-seventh aspects can be obtained, respectively, and the change point operation output signal is output at the rising change point of the input data signal. , Which includes information on the phase number of the clock signal constituting the N-phase clock signal, it is possible to provide a digital PLL circuit and a signal reproducing method that can more accurately select an extracted clock signal. .

According to the twelfth and twenty-ninth aspects of the present invention, the tenth and eleventh aspects and the second aspect, respectively, are provided.
Since the effect of the invention described in 7 or 28 is obtained, and the change point calculation output signal includes the information of the phase number of the clock signal constituting the N-phase clock signal indicating the falling change point of the input data signal, It is possible to provide a digital PLL circuit and a signal reproducing method that can more accurately select an extracted clock signal based on the information on the phase number.

According to the thirteenth and thirty aspects of the present invention, the tenth to twelfth and the second aspect, respectively, are provided.
The effect of the invention described in any one of 7 to 29 is obtained, and the change point calculation output signal includes information indicating the number of change points of the input data signal during one cycle of the extracted clock signal. Further, it is possible to provide a digital PLL circuit and a signal reproducing method capable of accurately reproducing a reproduced data signal.

According to the fourteenth and thirty-first aspects of the present invention, the tenth to the thirteenth aspects and the second aspect, respectively, are provided.
7 to 30 are obtained, and N sample data signals are fetched at a timing synchronized with the extracted clock signal, and the rising change point of the input data signal is determined from the fetched N values. A change point calculation output signal that includes the average value of the phase numbers with high accuracy because the calculation process that detects the phase number of the clock signal corresponding to the falling change point and calculates the average value of the detected phase numbers is performed. A digital PLL circuit and a signal reproducing method capable of outputting the same can be provided.

According to the invention of claims 15 and 32, claims 10 to 14 and claim 2 respectively.
The effect of the invention described in any one of 7 to 31 is obtained, and the average value of the phase numbers of the clock signal indicating the rising transition point of the input data signal, which is included in the transition point calculation output signal, or the input data signal A clock signal of the phase indicated by the average value of the phase number of the clock signal indicating the falling transition point is selected from the N-phase clock signals and output as the extracted clock signal, so that the rising transition point of the input data signal is obtained. Alternatively, it is possible to provide a digital PLL circuit and a signal reproducing method capable of outputting an extracted clock signal following a change in a falling transition point.

That is, the average value follows the phase (phase information) of the center value of the fluctuation (jitter) of the data change point at the rising or falling change point of the input data signal. The extracted clock signal selected according to the phase (phase information) indicated by the value is phase-locked and follows the input data signal, thereby absorbing the frequency deviation between the input data signal and the N-phase clock signal, and causing jitter and duty fluctuation. Digital PLL capable of reproducing a reproduced data signal without error with sufficient tolerance
A circuit and a signal reproducing method can be provided.

According to the invention of claims 16 and 33, claims 4 to 15 and 21 to 3 respectively.
2. The effects of the invention described in any one of 2) are obtained, and
Since the input N sample data signals are delayed while the phase difference between the sample data signals is maintained,
A digital PLL circuit and a signal reproducing method capable of further improving the convenience of a delayed sample data signal can be provided.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an embodiment of a digital PLL circuit according to the present invention.

FIG. 2 is a conceptual diagram of an operation when a change point detection calculation unit shown in FIG. 1 captures a value of a sample data signal.

FIG. 3 is a process chart of an operation when a change point detection calculation unit shown in FIG. 1 calculates an average value of phase numbers.

FIG. 4 is a circuit diagram showing a first configuration example of a delay processing unit shown in FIG. 1;

FIG. 5 is a circuit diagram showing a second configuration example of the delay processing unit shown in FIG. 1;

FIG. 6 is a time chart illustrating an operation of the data sampling processing unit illustrated in FIG. 1;

FIG. 7 is a time chart illustrating an operation of the delay processing unit illustrated in FIG. 1;

FIG. 8 is a time chart illustrating an operation of a change point detection calculation unit illustrated in FIG. 1;

FIG. 9 is a time chart illustrating an operation of an extracted clock selection processing unit illustrated in FIG. 1;

FIG. 10 is a time chart illustrating an operation of the data reproduction processing unit illustrated in FIG. 1;

FIG. 11 is a circuit diagram showing a configuration example of a data reproduction processing section shown in FIG. 1;

12 is a conceptual diagram showing a data reproduction operation of a data reproduction processing unit shown in FIG.

FIG. 13 is a diagram showing the number of bits of the pull-in in the present invention and the conventional digital PLL circuit and signal reproducing method.

FIG. 14 is a conceptual diagram showing a change point detection operation of an input data signal in the present invention and a conventional digital PLL circuit and signal reproducing method.

FIG. 15 is a conceptual diagram of a data reproducing method according to the present invention and a conventional digital PLL circuit and signal reproducing method.

FIG. 16 is a block diagram showing a configuration of a conventional digital PLL circuit.

FIG. 17 is a block diagram showing a configuration of a conventional digital PLL circuit.

[Explanation of symbols]

Reference Signs List 1 data sampling processing unit 2 delay processing unit 3 data reproduction processing unit 4 change point detection calculation unit 5 extracted clock signal selection processing unit 6 sample data signal (N lines) 7 delayed sample data signal (N lines) 8 change point calculation output signal 11-1-1, 11-1-2,..., 11-1-M
Flip-flops 11-2-1, 11-2-2, ..., 11-2-M
Flip-flops 11-N-1, 11-N-2, ..., 11-N-M
Flip-flop 12 L frequency dividing circuit 201 Subtractor 202 Weighting unit 203 Adder 204 Averaging unit 801 Selection circuit 802 Encoder 804 Selection circuit 805 Flip-flop C0 0th phase clock signal C1 1st phase clock signal C2 2nd Phase clock signal C3 third phase clock signal C4 fourth phase clock signal C5 fifth phase clock signal C6 sixth phase clock signal C7 seventh phase clock signal D0 zero phase sample data signal D1 first Phase sample data signal D2 second phase sample data signal D3 third phase sample data signal D4 fourth phase sample data signal D5 fifth phase sample data signal D6 sixth phase sample data signal D7 seventh phase Sample data signal R0 Delay of phase 0 Sample data signal R1 Delay of phase 1 Sample data signal R2 second phase delayed sample data signal R3 third phase delayed sample data signal R4 fourth phase delayed sample data signal R5 fifth phase delayed sample data signal R6 sixth phase delayed sample data signal R7 7-phase delayed sample data signal

Claims (33)

[Claims] 1. A digital PLL circuit for outputting a reproduced data signal synchronized with an extracted clock signal, wherein a sampled data signal is delayed to output a reproduced data signal synchronized with the extracted clock signal at high speed and with a small error. A digital PLL circuit characterized by the above-mentioned. 2. The method according to claim 1, wherein the sampling data signal obtained by sampling the input data signal with an N-phase clock signal is delayed, and the time until the extracted clock signal is selected is the time until the reproduced data signal is reproduced. A digital PLL circuit which outputs a reproduced data signal synchronized with the extracted clock signal at high speed with a small error by the following. 3. A digital system having at least a data sampling processor for sampling an input data signal and outputting N sample data signals, and a data reproduction processor for outputting a reproduced data signal synchronized with an extracted clock signal. In the PLL circuit, a delay processing unit that delays the N sample data signals is provided between the data sampling processing unit and the data reproduction processing unit, and a time until the extracted clock signal is selected is set. A digital PLL circuit for outputting a reproduced data signal synchronized with the extracted clock signal at high speed and with a small error by setting the time to be equal to or less than the time until the reproduced data signal is reproduced. 4. An input data signal and an N-phase clock signal composed of N clock signals are input, and the input data signal is sampled by the N-phase clock signal to generate N sample data signals. And a data reproduction processing unit that outputs a reproduced data signal synchronized with the extracted clock signal. An extraction synchronized with the reproduced data signal from the N-phase clock signal Digital P that outputs clock signal
In the LL circuit, the N sample data signals output from the data sampling processing unit are input, the input N sample data signals are delayed, and the N data signals are processed as the N delayed sample data signals. And a time until the extracted clock signal is selected and output based on the input data signal is equal to or less than a time until the input data signal is reproduced as a reproduction data signal. A digital PLL circuit that outputs a reproduced data signal synchronized with the extracted clock signal at high speed and with a small error. 5. The N-phase clock signal has substantially the same frequency as the input data signal and has a phase of 36.
5. The digital PLL circuit according to claim 4, wherein the digital PLL circuit is a signal composed of N clock signals sequentially shifted by 0 degrees / N (N is an integer of 2 or more). 6. The data reproduction processing unit selects one delay sample data signal from the N delay sample data signals output from the delay processing unit, and reproduces the data in synchronization with an extracted clock signal. The digital PLL circuit according to claim 4, wherein the digital PLL circuit outputs the data as a data signal. 7. The digital PLL circuit according to claim 4, further comprising an extracted clock selection processing unit that selects and outputs the extracted clock signal from the N-phase clock signals. . 8. The delay processing unit has N flip-flop lines, each of which is constituted by M stages (M is an arbitrary integer of 1 or more) of flip-flops to which N sample data signals are input. 8. The digital PLL circuit according to claim 4, wherein a clock signal input to a flip-flop forming each flip-flop line is a clock signal forming the N-phase clock signal. 9. The delay processing unit comprises: N flip-flop lines, each of which is constituted by M (M is an integer of 1 or more) flip-flops to which N sample data signals are input; And an L frequency dividing circuit that divides each of the N-phase clock signals into 1 / L (L is an arbitrary integer of 2 or more). The clock signal input to the flip-flops constituting each flip-flop line is 8. The digital PLL circuit according to claim 4, wherein the clock signal forming the N-phase clock signal is a signal obtained by dividing the frequency by the L divider circuit. 10. The N sampling data signals output from the data sampling processing section and the extracted clock signal are input, and the extraction is performed based on the input N sample data signals and the extracted clock signal. 10. The digital device according to claim 4, further comprising: a clock selection processing unit; and a change point detection calculation unit that outputs a signal for controlling the data reproduction processing unit as a change point calculation output signal. PLL circuit. 11. The N signal, wherein the change point calculation output signal indicates a falling change point of the input data signal.
11. The digital P according to claim 10, further comprising information of a phase number of a clock signal constituting the phase clock signal.
LL circuit. 12. The N signal, wherein the change point calculation output signal indicates a rising change point of the input data signal.
12. The digital PLL circuit according to claim 10, further comprising information on a phase number of a clock signal constituting the phase clock signal. 13. The digital PLL circuit according to claim 10, wherein the change point calculation output signal includes information indicating the number of change points of the input data signal. 14. The change point detecting and calculating section fetches the N sample data signals at a timing synchronized with the extracted clock signal, and calculates a rising change point of the input data signal from the fetched N values. The change point calculation output signal is output by detecting a phase number of a clock signal corresponding to a falling change point and performing an arithmetic process of calculating an average value of the detected phase numbers. Item 14. Digital PLL according to any one of Items 10 to 13
circuit. 15. The extracted clock selection processing unit includes: a phase number of a clock signal indicating a rising transition point of the input data signal, the phase number being included in the transition point computation output signal output from the transition point detection computation unit; An average value or a clock signal of a phase indicated by the average value of the phase number of the clock signal indicating the falling transition point of the input data signal is selected from the N-phase clock signals and output as an extracted clock signal. 15. The digital PLL circuit according to claim 10, wherein the digital PLL circuit outputs an extracted clock signal that follows a change in a rising transition point or a falling transition point of the input data signal. 16. The method according to claim 4, wherein the delay processing section delays the input N sample data signals while maintaining a phase difference between the N sample data signals. The digital PLL circuit according to any one of the above. 17. An N clock signal comprising an input data signal and N clock signals having substantially the same frequency as the input data signal and having a phase shifted sequentially by 360 degrees / N (N is an integer of 2 or more). And the input data signal is sampled by the N-phase clock signal.
A data sampling processor (1) for outputting N sample data signals; N sample data signals (6) output from the data sampling processor; and the N-phase clock signal; A delay processing unit (2) for delaying the N sampled data signals based on the N-phase clock signal and outputting as N delayed sampled data signals (7);
And the N sampled data signals output from the data sampling processing unit and the extracted clock signal output from the extracted clock selection processing unit (5) are input, and the extracted clock selection processing unit and data reproduction are input. A change point detection calculation section (4) for outputting a change point calculation output signal (8) for controlling the processing section (3); and the N-phase clock signal is input and output from the change point detection calculation section. Based on the change point calculation output signal, a clock signal to be output as an extracted clock signal is selected from the N-phase clock signal, and this is used as the extracted clock signal to output the change point detection calculation unit, data reproduction processing unit, and extraction to the outside A clock selection processing unit (5); the N delayed sample data signals output from the delay processing unit; and a change point output from the change point detection calculation unit And calculating power signal, said extracted output from the clock selection unit extraction clock signal are inputted, a digital PLL circuit; and a data reproduction processing section for outputting the reproduced data signal (3). 18. A signal reproducing method for outputting a reproduced data signal synchronized with an extracted clock signal, the method comprising: outputting a reproduced data signal synchronized with the extracted clock at a high speed with a small error by delaying a sample data signal. Characteristic signal reproduction method. 19. A sample data signal obtained by sampling an input data signal with an N-phase clock signal is delayed, and a time until an extracted clock signal is selected is a time until a reproduced data signal is reproduced. A signal reproducing method comprising: outputting a reproduced data signal synchronized with the extracted clock signal at high speed with a small error. 20. A signal having at least a data sampling processing step of sampling an input data signal and outputting N sample data signals, and a data reproduction processing step of outputting a reproduction data signal synchronized with an extracted clock signal. In the reproduction method, a delay processing step of delaying the N sample data signals is provided between the data sampling processing step and the data reproduction processing step, and a time until the extracted clock signal is selected is provided. A signal reproducing method comprising: outputting a reproduced data signal synchronized with the extracted clock signal at high speed and with a small error by setting the reproduction data signal to a time equal to or less than the time required for reproducing the reproduced data signal. 21. An input data signal and an N-phase clock signal composed of N clock signals are inputted,
A data sampling process of sampling the input data signal by a phase clock signal and outputting N sample data signals; and a data reproduction process of outputting a reproduced data signal synchronized with the extracted clock signal. A signal reproducing method for outputting an extracted clock signal synchronized with the reproduced data signal from the N-phase clock signals, wherein the N sample data signals output from the data sampling processing step are input; A delay processing step of delaying the input N sample data signals and outputting them as N delayed sample data signals; and a time until an extracted clock signal is selected and output based on the input data signal is: By setting the input data signal to be equal to or less than the time until it is selected as a reproduction data signal, Signal reproducing method and outputting a reproduced data signal synchronized with the serial extraction clock signal at a high speed and with fewer errors. 22. The N-phase clock signal has a frequency substantially equal to that of the input data signal and a phase of 36.
22. The signal reproducing method according to claim 21, wherein the signal is an N-phase clock signal by N clock signals sequentially shifted by 0 degrees / N (N is an integer of 2 or more). 23. The data reproduction processing step, wherein one delay sample data signal is selected from the N delay sample data signals output in the delay processing step and reproduced in synchronization with an extracted clock signal. 22. The signal output as a data signal.
3. The signal reproducing method according to 2. 24. An extraction clock selection processing step of selecting and outputting the extracted clock signal from the N-phase clock signal.
The signal reproducing method according to any one of the above. 25. The delay processing step, wherein the N sample data signals are each constituted by M stages (M is an arbitrary integer of 1 or more) of flip-flops,
The input N sample data signals are input to the N flip-flop lines, and the clock signals forming the N-phase clock signal are used as the clock signals input to the flip-flops forming the respective flip-flop lines. 25. The signal reproducing method according to claim 21, which is a step of delaying. 26. The delay processing step, wherein each of the N sample data signals is constituted by M stages (M is an arbitrary integer of 1 or more) of flip-flops,
Each of the input N-phase clock signals is input to an N frequency-dividing circuit that divides the input N-phase clock signal by 1 / L (L is an arbitrary integer of 2 or more). Delaying the input N sample data signals by using, as a clock signal to be input to the flip-flop, a signal obtained by dividing the clock signal forming the N-phase clock signal by the L divider circuit The signal reproducing method according to any one of claims 21 to 24, wherein: 27. A signal for controlling the extraction clock selection processing step and the data reproduction processing step based on the N sample data signals output in the data sampling processing step and the extraction clock signal. 27. The signal reproducing method according to claim 21, further comprising a change point detection calculation step of outputting as a change point calculation output signal. 28. The N signal, wherein the change point calculation output signal indicates a falling change point of the input data signal.
28. The signal reproducing method according to claim 27, further comprising information on a phase number of a clock signal constituting the phase clock signal. 29. The N signal, wherein the change point calculation output signal indicates a rising change point of the input data signal.
29. The signal reproducing method according to claim 27, further comprising information on a phase number of a clock signal constituting the phase clock signal. 30. The signal reproducing method according to claim 27, wherein the change point calculation output signal includes information indicating the number of change points of the input data signal. 31. The change point detecting operation step: fetching the N sample data signals at a timing synchronized with the extracted clock signal, and calculating a rising change point of the input data signal from the fetched N values. The change point calculation output signal is output by detecting a phase number of a clock signal corresponding to a falling change point and performing an arithmetic process of calculating an average value of the detected phase numbers. Item 31. The signal reproducing method according to any one of Items 27 to 30. 32. A phase number of a clock signal indicating a rising change point of the input data signal, wherein the extracted clock selection processing step is included in the change point calculation output signal output in the change point detection calculation step. An average value or a clock signal of a phase indicated by the average value of the phase number of the clock signal indicating the falling transition point of the input data signal is selected from the N-phase clock signals and output as an extracted clock signal. 32. The signal reproducing method according to claim 27, wherein the extracted clock signal is output so as to follow a change of a rising transition point or a falling transition point of the input data signal. 33. The method according to claim 21, wherein the delay processing step delays the input N sample data signals while maintaining a phase difference between the N sample data signals. The signal reproducing method according to any one of the above.